Thanks for all your comments – I'll come back to this once I finished and still haven't run out of puff.
The first part of the "HowTo - as simple as it gets - for DIY CSD measurement" was about what CSD is good for and about some limitations applying.
It can be found here:
http://www.diyaudio.com/forums/showthread.php?postid=1505009#post1505009
The second part of the "HowTo - as simple as it gets - for DIY CSD measurement" was about the exceptional low setup requirements and the fairly cheap equipment needed to capture useful CSD's.
It can be found here:
http://www.diyaudio.com/forums/showthread.php?postid=1506069#post1506069
Once you have set up your speaker and mic – plugged all the cables and booted your PC (which may turn out to be the most time consuming action for the whole CSD measurement) you are ready to start measuring right away.
You can use the default settings of ARTA by no problem.
Press "Record" >> "Impulse Response" and you will see the popup window below.
Chose "MLS" or "Periodic noise" whichever you like more:
Try to press "Record" at a low level of your volume control.
If you get sound - adjust it fairly loud but not overloading the speaker. Now adjust the mic gain until you get the bar graph somewhere close to –10 / -20dB.
If something does not work - check your cables, your soundcard and spend some time with the ARTA manual or post in news group - DON'T ASK ME - I'm just a monkey pressing buttons like others throwing up letters...
Same to all the options available to fine tune or magnify certain aspects - DON'T ASK ME - I'm just the monkey ...waiting to see the letters falling down as a poem of Hesse
Well lets assume you got some signal to your speaker and some signal back from your mic - now you should see an impulse response like below.
If you now press "CSD" you will get something like below
confirm the settings with "OK" and you are there
Well looks pretty good for a starting point, eh?
The data prior to decay isn't that interesting here so we go back to the impulse response and set a roughly 5ms time window ("Gate") with the left and right mouse button like below :
The CSD we got already looks great.
Wow TWO prominent resonance - the lower one obviously "walking down by steps" – relatively clean operation below and above that.
What we can further explore is what we see when setting the window ("Gate") - just for fun - somewhere at the IR like below:
Not much difference – just a little later in time the characteristic of that resonance still is the same.- the lower frequencies are almost dead at this time though.
What happens if we use longer window - roughly 13ms - can be seen below
In short- not that much more detail to gain
Way more interesting is what we see if we set the window ("Gate") back to our 5ms and increase the Z-Axis of the CSD plot to 70dB as is shown below
- first and most obvious - ARTA crops the display where no valid data is available (> 5ms / < 250Hz)
- second thing we see is that there are steps for both resonant freqeuncies
- and third we can see something that looks like kind of bungee jumping between the two prominent resonance's.
This is better seen in the Sonogram where deep holes alternate with bridges between the two resonance's.
Note that in these CSD's we didn't see much penalty even if we intentionally look at and "Gate" the IR at a time where full bottom / ceiling reflection occurs.
Test-question – what do we see below?
YES ! Right ! exactly !
Its my neighbours water pump at 2kHz – if I am not completely wrong.
Luckily we don't hear that pump – the sound is completely buried in the noise spectra.
Left hand we see the beauty of mild wind in May and ours city sound at evening time.
Greetings
Michael
The first part of the "HowTo - as simple as it gets - for DIY CSD measurement" was about what CSD is good for and about some limitations applying.
It can be found here:
http://www.diyaudio.com/forums/showthread.php?postid=1505009#post1505009
The second part of the "HowTo - as simple as it gets - for DIY CSD measurement" was about the exceptional low setup requirements and the fairly cheap equipment needed to capture useful CSD's.
It can be found here:
http://www.diyaudio.com/forums/showthread.php?postid=1506069#post1506069
Once you have set up your speaker and mic – plugged all the cables and booted your PC (which may turn out to be the most time consuming action for the whole CSD measurement) you are ready to start measuring right away.
You can use the default settings of ARTA by no problem.
Press "Record" >> "Impulse Response" and you will see the popup window below.
Chose "MLS" or "Periodic noise" whichever you like more:
An externally hosted image should be here but it was not working when we last tested it.
Try to press "Record" at a low level of your volume control.
If you get sound - adjust it fairly loud but not overloading the speaker. Now adjust the mic gain until you get the bar graph somewhere close to –10 / -20dB.
If something does not work - check your cables, your soundcard and spend some time with the ARTA manual or post in news group - DON'T ASK ME - I'm just a monkey pressing buttons like others throwing up letters...
Same to all the options available to fine tune or magnify certain aspects - DON'T ASK ME - I'm just the monkey ...waiting to see the letters falling down as a poem of Hesse
Well lets assume you got some signal to your speaker and some signal back from your mic - now you should see an impulse response like below.
An externally hosted image should be here but it was not working when we last tested it.
If you now press "CSD" you will get something like below
An externally hosted image should be here but it was not working when we last tested it.
confirm the settings with "OK" and you are there
An externally hosted image should be here but it was not working when we last tested it.
Well looks pretty good for a starting point, eh?
The data prior to decay isn't that interesting here so we go back to the impulse response and set a roughly 5ms time window ("Gate") with the left and right mouse button like below :
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The CSD we got already looks great.
Wow TWO prominent resonance - the lower one obviously "walking down by steps" – relatively clean operation below and above that.
What we can further explore is what we see when setting the window ("Gate") - just for fun - somewhere at the IR like below:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Not much difference – just a little later in time the characteristic of that resonance still is the same.- the lower frequencies are almost dead at this time though.
What happens if we use longer window - roughly 13ms - can be seen below
An externally hosted image should be here but it was not working when we last tested it.
In short- not that much more detail to gain
Way more interesting is what we see if we set the window ("Gate") back to our 5ms and increase the Z-Axis of the CSD plot to 70dB as is shown below
An externally hosted image should be here but it was not working when we last tested it.
- first and most obvious - ARTA crops the display where no valid data is available (> 5ms / < 250Hz)
- second thing we see is that there are steps for both resonant freqeuncies
- and third we can see something that looks like kind of bungee jumping between the two prominent resonance's.
An externally hosted image should be here but it was not working when we last tested it.
This is better seen in the Sonogram where deep holes alternate with bridges between the two resonance's.
Note that in these CSD's we didn't see much penalty even if we intentionally look at and "Gate" the IR at a time where full bottom / ceiling reflection occurs.
Test-question – what do we see below?
An externally hosted image should be here but it was not working when we last tested it.
YES ! Right ! exactly !
Its my neighbours water pump at 2kHz – if I am not completely wrong.
Luckily we don't hear that pump – the sound is completely buried in the noise spectra.
Left hand we see the beauty of mild wind in May and ours city sound at evening time.
Greetings
Michael
One thing I forgot to outline that is also too obvious in the sonogram picture of my last posting.
There is a structure I would like to describe as the "Christmas tree".
The resonance's spread out to lower and higher frequencies symmetrically. With this driver (roughly the size of the B110 and also measured "naked" at around 10cm mic distance) it correlates with the bungee holes but it also can be observed at drivers with a single resonance.
Kind of concentrating energy to a smaller frequency band or broadening the spectrum in intervals during decay.
Greetings
Michael
There is a structure I would like to describe as the "Christmas tree".
The resonance's spread out to lower and higher frequencies symmetrically. With this driver (roughly the size of the B110 and also measured "naked" at around 10cm mic distance) it correlates with the bungee holes but it also can be observed at drivers with a single resonance.
Kind of concentrating energy to a smaller frequency band or broadening the spectrum in intervals during decay.
Greetings
Michael
gedlee said:
No insult intended. I said the best minds have left audio, not that Floyde Toole, Dick Small, or you were incompetent. I don't think you'll find the word "incompetent" anywhere in the original posting, nor a single mention of FT, DS, or you. I have no idea why you are taking personally a posting directed at the state of the entire audio industry.
The sad fact is that Dr. Amar Bose collects Ph.D. grads as trophies, and gives them nothing to do - it doesn't take an Einstein to design home-theater-in-a-box with Shenzen-made drivers. I'm not aware of Harman International, TAD, THX, Wilson Audio, Audio Research, or anyone I can think of hiring anyone on the level of Hilliard, Blumlein, or Major Armstrong. Do we have any Armstrongs today? No. Do we have any Blumleins? No.
There are plenty of people at this level in other fields, but audio isn't one of them. We've had a few greats lately, but they're getting thinner on the ground over the decades, not more plentiful. The trend over the last seventy years is unmistakable.
From the outside, audio is considered a dull field, with all the important problems solved, so we get the dregs of engineering, the "value engineers", figuring out ways to de-content the product and relentlessly drive down costs. That's not all of the engineers, thank goodness, but look where the serious R&D money is going. Compression, cell phones, iPods, and lots of military spending on underwater acoustics.
I left the AES when most of the papers started to be about "inaudible" lossy-compression schemes - never mind whether it is "inaudible" or not, the entire purpose of compression is to save money - at a time when transmission and storage costs are relentlessly following Moore's Law. Why should the best and brightest be interested when much of the R&D money in the industry is dedicated to squeezing a quart into a pint pot? Who cares? And what does any of that have to with improving quality?
I don't know who is responsible for this state of affairs. Consumers have lost the taste for live, acoustic music - that's considered, in current American parlance, the deadliest insult of all - "elitist". Easy enough to sling at someone you don't like, but what does it actually mean? Someone who doesn't have your tastes, or who doesn't live in your part of the country? It must be some kind of American culture-war thing, and you need a decoder ring to figure out what's elitist and what's not.
Audiophiles, previously known as hi-fi nuts? The demographics of this group haven't changed in fifty years - the customer base is still doctors, lawyers, entrepreneurs, successful engineers, and almost entirely men in their thirties and older. No change there at all, except that the demographic is getting older - and is probably the same cohort, which does not bode well for the future.
The manufacturers? Hmm, hard to say. The Harvard biz-school types figured out by the early Seventies that marketing, and stock-price manipulation, had a much better payoff than R&D. Thus the steady de-industrialization of the country, and rise of the Wall Street "service" sector. The hifi industry has been a minor bystander in this whole process. Good-bye RCA, Altec, Bell Labs, Marantz, Scott, Fisher, Sherwood et al. It was nice knowing you.
The American audio-review magazines? Well, yes, now we have a bone to pick. The trend of corruption over the last thirty years has been unmistakable. This is an open secret in the industry, and only the public doesn't know - but many former readers, if not most, have their suspicions.
All of these trends have been going on for decades, and are part of much larger social changes in this country and across the world. We are bystanders to the immense changes of our time. But still ... we can create a little tiny space for excellence, even if the larger culture thinks we're doing nothing more interesting than playing with model trains. Our invisibility might actually be an asset, giving us a little more room to move.
I want to thank mige0 for the wonderfully clear and lucid postings. This is an area that is commonly misunderstood, and the sloppy measurements in the magazines don't help any.
I for one am very grateful that all of us, with only a modest expenditure on our PCs, can make quite detailed and revealing measurements of commercial systems and our own DIY efforts. The day of quick-n-dirty RTA single and third-octave measurements is finally over.
Thank you, D.E.L. Shorter, Richard Heyser, S.P. Lipshitz, and John Vanderkooy. We stand on the shoulders of giants.
I for one am very grateful that all of us, with only a modest expenditure on our PCs, can make quite detailed and revealing measurements of commercial systems and our own DIY efforts. The day of quick-n-dirty RTA single and third-octave measurements is finally over.
Thank you, D.E.L. Shorter, Richard Heyser, S.P. Lipshitz, and John Vanderkooy. We stand on the shoulders of giants.
gedlee said:
but why? people LOVE music! and listen to it at home
there is certainly a market for "a sound quality"
people attend the shows, sales continue, the prices get higher and higher and so on, there is even renewed interest in stereo systems
so why is there no market for GOOD sound quality?
do the people prefer bad sound?
or perhaps not the people but just "audiophiles"?
and the rest of the potential market ie. non-audiophile music lovers are simply repelled by the "audiophile sound" combined with insane pricing and all this weird irrational gadget-concentrated "audiophile" shopaholic loner subculture?
two problems:
1) inaccurate sound which even a person with impaired hearing can judge...
and I agree with Linkwitz on that:
everyone can tell this "audiophile" big buck sound form the real thing!
so what is it all about?
"living in the age of nuclear physics and so on" and all we are able to achieve is such a pitiful flat imitation?
isn't it nonsense?
be serious - there is no breakthrough in realism of audio reproduction between decent minisystem and many big buck audiophile stereo systems
this is why it is "much ado about nothing" for most people
2) ...combined with wierd/silly social image of an "audiophile" (mains cables claims and the rest)
"an audiophile(...) seems to want to reproduce the sound of a triangle more triangular than the real thing" in the words of H.A. Hartley
and the same audiophile accepts at the same time that pitiful flat imitation of the real thing and pays big bucks allegedly for that...
...and in reality for mains cable and so on
this is silly, weird, even disturbing for most people
IMO these are the real impediments to the HiFi market growth
dissapointing unrealistic sound quality obvious to anyone with normal (ie. non-audiophile) hearing
and deformed image of a HiFi interested person created by years of niche marketing, of breeding of loyal consumers willing to buy all those gadgets and obsessively returning to buy again
buy new and newer and more and more expensive gear
constantly switching from one to another and even back and again
things like "one cd player for pop music and another for classical" for example
pure nonsense
a shopaholic - an ideal consumer
best,
graaf
That's funny, just the opposite is true, in terms of audio reproduction our capacity is still in the "infant" stage as far as I'm concerned. ALL speakers basically suck. There is just a wide range of suckyness from mediocre to terrible... Some of them may be pleasurable to listen to, but, by no means, are they capable of a realistic reproduction of all ranges of music. Some speakers are very good at some parts of music production, but they all have flaws, compromises, areas where they fall short.Lynn Olson said:
graaf said:
Having sat in some interesting meetings and demonstrations with very influential people in the much larger realm of consumer electronics, it's most certainly clear there is a market for sound quality...
The hurdle is to package and present it in a way that people will buy it.
Consumers really do want it! They just don't want it to be more difficult or complicated...
Mark Seaton said:
You are absolutely right!
Mark Seaton said:
...or "room unfriendly" that is demanding special decicated room (for "another beloved family member") to sing
otherwise completely ruining (when properly set up) the arrangement and decor of normal living room
or just silly like "audiophile" cables
and so on
most people feel uncomfortable doing silly things like buying something like that
best,
graaf
Hello,
Well, up untill now it seems to me that most of the band is properly covered from a technologic point of view.
1. We have a high SPL ribbon from 6-7Khz and up to 30Khz with excelent transients and time-domain performance.
2. We have 1.4' compression drivers with aluminium or composite diaphragm loaded into OS or Le'cleach waveguides (short horns), with excelent power-handling, excellent disspersion pattern and very good time domain performance if the horn is designed properly (and if it's kept short?). And all this avoiding break-up resonances. This would be the 800-900Hz - 6-7Khz band.
3. Now, for the 50Hz-300Hz, we can use large bass arrays in open baffle. The configuration of the array (4*4 for example) would allow us even to control the dispersion pattern. Considering that high SPL 12' and 15' are usual in the industry it has the advanatge of being accesible. The preference for low Qts drivers in OB makes it affordable. And all this while retaining some really good performance as for example Magnetar and other memebers of this forum demonstrated.
4. Now, for me the biggest concern is the so called power band or music band or whatever do you want to call it. And it's roughly in the 200Hz-1Khz region. From my point of view, there is now technology (at least an affordable one) to keep with the performance of the rest.
Large arrays of direct radiators probably are flawed as comb effects and other errors can appear because of the relativelly high frequency.
Single direct radiators don't have the spl capabilities in OB to keep up with the rest. Besides that, 12" and 15" drivers may have resonances in the upper part of this band.
Horns tend to be large in this region and at 1Khz crossover point you can end up with a distance between acoustic centers of 90cm making crossover design a hell. But that's not the biggest problem. I know no compression driver to have the same distorsion performances in this band as a direct radiator. (Maybe some exotic ones like the 6" emilar, goto, ale or he 4'' community driver).
Sticking a high-efficiency direct radiator in a waveguide seems a solution but specialists in horn theory like Dr. Geddes disaprove this because of the problems at the throat of the horn and wavefront form.
Now.. am I missing something? Or maybe I am concerning too much?
Well, up untill now it seems to me that most of the band is properly covered from a technologic point of view.
1. We have a high SPL ribbon from 6-7Khz and up to 30Khz with excelent transients and time-domain performance.
2. We have 1.4' compression drivers with aluminium or composite diaphragm loaded into OS or Le'cleach waveguides (short horns), with excelent power-handling, excellent disspersion pattern and very good time domain performance if the horn is designed properly (and if it's kept short?). And all this avoiding break-up resonances. This would be the 800-900Hz - 6-7Khz band.
3. Now, for the 50Hz-300Hz, we can use large bass arrays in open baffle. The configuration of the array (4*4 for example) would allow us even to control the dispersion pattern. Considering that high SPL 12' and 15' are usual in the industry it has the advanatge of being accesible. The preference for low Qts drivers in OB makes it affordable. And all this while retaining some really good performance as for example Magnetar and other memebers of this forum demonstrated.
4. Now, for me the biggest concern is the so called power band or music band or whatever do you want to call it. And it's roughly in the 200Hz-1Khz region. From my point of view, there is now technology (at least an affordable one) to keep with the performance of the rest.
Large arrays of direct radiators probably are flawed as comb effects and other errors can appear because of the relativelly high frequency.
Single direct radiators don't have the spl capabilities in OB to keep up with the rest. Besides that, 12" and 15" drivers may have resonances in the upper part of this band.
Horns tend to be large in this region and at 1Khz crossover point you can end up with a distance between acoustic centers of 90cm making crossover design a hell. But that's not the biggest problem. I know no compression driver to have the same distorsion performances in this band as a direct radiator. (Maybe some exotic ones like the 6" emilar, goto, ale or he 4'' community driver).
Sticking a high-efficiency direct radiator in a waveguide seems a solution but specialists in horn theory like Dr. Geddes disaprove this because of the problems at the throat of the horn and wavefront form.
Now.. am I missing something? Or maybe I am concerning too much?
low midrange
I have to say I work on that area (150 to 1K) a lot - it is the core of the music, and either a big motor 6-8" low mass driver in a front horn (300-1.8k) or a deluxe alnico 12 that will go an octave lower (Like the custom EV SRO's I've got hummin right now) with a tad of eq or big baffle to get them down to 120-150 cycles is the way to go. The 12 is easier and can have nearly the same sensitivity as the horn loaded cone, problem is it seems no one currently makes one that that i'd consider - maybe the tone tubby but they seem a bit weak in the magnet
I have to say I work on that area (150 to 1K) a lot - it is the core of the music, and either a big motor 6-8" low mass driver in a front horn (300-1.8k) or a deluxe alnico 12 that will go an octave lower (Like the custom EV SRO's I've got hummin right now) with a tad of eq or big baffle to get them down to 120-150 cycles is the way to go. The 12 is easier and can have nearly the same sensitivity as the horn loaded cone, problem is it seems no one currently makes one that that i'd consider - maybe the tone tubby but they seem a bit weak in the magnet
Well, I doubt single speakers can do it. Maybe in small rooms, on low sound levels and with certain music genres.
I agree however that it's easier to talk about coherence with wide-range drivers. I don't have the experience to talk about if it's possible to get the sensation of coherence with multiway systems.
SunRa said:
I have Arta now and have measured around 15 different low mid / compression driver combo systems in about every way the software is capable of (I'm getting it) - if you want to lower the distortion on a low mid driver use an electronic 4th order crossover at 2 to 3 times the fs of the driver. Same goes for both small and large format compression drivers. Once the distortion is below 1 percent at 110 db (yes 110 only use the best drivers you can find) there is a large awakening of sorts. It becomes more coherent and 'free' - don't 'worry' about the steep slopes, they sound best - This is something I have experienced for a long time now but am now correlating it to measurements since it's so easy to do.
Re: low midrange
How about what Lynn was considering (and I am scratching my head on the right way to get it right): Namely use two matched drivers for that band, with "proper" driver summation. For example the above 12NDA520 + 6ND410 combo, sharing a LR2 at 200 Hz and 1.5k respectively, with 12NDA520 rolled off (first order ?) on the last top octave and the 6ND410 HP brought in (again, first order ?) one octave later down below...
Magnetar said:
How about what Lynn was considering (and I am scratching my head on the right way to get it right): Namely use two matched drivers for that band, with "proper" driver summation. For example the above 12NDA520 + 6ND410 combo, sharing a LR2 at 200 Hz and 1.5k respectively, with 12NDA520 rolled off (first order ?) on the last top octave and the 6ND410 HP brought in (again, first order ?) one octave later down below...